Recurrent neural network-based volumetric fluorescence microscopy.
Recurrent neural network-based volumetric fluorescence microscopy.
复制标题
DOI：
10.1038/s41377-021-00506-9
复制
发表时间：
2021-03-23
期刊：
影响因子：
--
通讯作者：
Ozcan A
中科院分区：
文献类型：
--
作者：
Huang L;Chen H;Luo Y;Rivenson Y;Ozcan A
利用荧光显微镜对样品进行体积成像在包括物理、医学和生命科学在内的各个领域发挥着重要作用。在这里，我们报告了一个基于深度学习的体积图像推理框架，该框架使用由标准宽视场荧光显微镜在样品体积内任意轴向位置稀疏捕获的2D图像。通过循环卷积神经网络，我们称之为recurrent - mz，从样品内的几个轴向平面的二维荧光信息被明确合并，以在扩展的视场深度上数字化重建样品体积。通过对秀丽隐杆线虫和纳米颗粒样品的实验，Recurrent-MZ被证明可以显著增加63x /1.4NA物镜的景深，并且可以将成像相同样品体积所需的轴向扫描次数减少30倍。我们通过展示其对不同成像条件（例如，不同的输入图像序列，涵盖各种轴向排列和未知轴向定位误差）的弹性，进一步说明了该循环网络在3D成像中的泛化。我们还演示了使用recursion - mz框架进行宽视场到共焦交叉模态图像转换，并使用一些宽视场2D荧光图像作为输入对样品进行3D图像重建，以匹配相同样品体积的共焦显微镜图像。recursion - mz展示了递归神经网络在显微图像重建中的首次应用，并提供了一个灵活快速的体积成像框架，克服了当前3D扫描显微镜工具的局限性。
Volumetric imaging of samples using fluorescence microscopy plays an important role in various fields including physical, medical and life sciences. Here we report a deep learning-based volumetric image inference framework that uses 2D images that are sparsely captured by a standard wide-field fluorescence microscope at arbitrary axial positions within the sample volume. Through a recurrent convolutional neural network, which we term as Recurrent-MZ, 2D fluorescence information from a few axial planes within the sample is explicitly incorporated to digitally reconstruct the sample volume over an extended depth-of-field. Using experiments on C. elegans and nanobead samples, Recurrent-MZ is demonstrated to significantly increase the depth-of-field of a 63×/1.4NA objective lens, also providing a 30-fold reduction in the number of axial scans required to image the same sample volume. We further illustrated the generalization of this recurrent network for 3D imaging by showing its resilience to varying imaging conditions, including e.g., different sequences of input images, covering various axial permutations and unknown axial positioning errors. We also demonstrated wide-field to confocal cross-modality image transformations using Recurrent-MZ framework and performed 3D image reconstruction of a sample using a few wide-field 2D fluorescence images as input, matching confocal microscopy images of the same sample volume. Recurrent-MZ demonstrates the first application of recurrent neural networks in microscopic image reconstruction and provides a flexible and rapid volumetric imaging framework, overcoming the limitations of current 3D scanning microscopy tools.
登录
查看更多内容
影响因子：
48
作者：
Abrahamsson, Sara;Chen, Jiji;Hajj, Bassam;Stallinga, Sjoerd;Katsov, Alexander Y.;Wisniewski, Jan;Mizuguchi, Gaku;Soule, Pierre;Mueller, Florian;Darzacq, Claire Dugast;Darzacq, Xavier;Wu, Carl;Bargmann, Cornelia I.;Agard, David A.;Dahan, Maxime;Gustafsson, Mats G. L.
通讯作者：
Gustafsson, Mats G. L.
影响因子：
10.6
作者：
Fang L;Li S;McNabb RP;Nie Q;Kuo AN;Toth CA;Izatt JA;Farsiu S
通讯作者：
Farsiu S
影响因子：
48
作者：
Chhetri, Raghav K.;Amat, Fernando;Kellerbeams, Philipp J.
通讯作者：
Kellerbeams, Philipp J.
影响因子：
10.4
作者：
Badon A;Bensussen S;Gritton HJ;Awal MR;Gabel CV;Han X;Mertz J
通讯作者：
Mertz J
影响因子：
14.8
作者：
通讯作者：
--